Last week, Effi­cient Fron­tier reported a 40% QoQ increase in Face­book CPCs, and this week TBG reported an approx­i­mately 30% drop in their U.S. cam­paigns. As Twit­ter chat­ter and the blo­gos­phere debate the results, there are a few key dif­fer­ences in the num­bers that have been miss­ing from the discussion.

First, we reported on the actual CPCs that mar­keters paid while TBGs num­bers reported trends in the sug­gested min and max CPCs on Face­book.  There­fore, the dif­fer­ence in per­cent­age growth and decline  are an apples to oranges com­par­i­son.  Sug­gested CPCs are a func­tion of Facebook’s algo­rithms and less a func­tion of what adver­tis­ers are will­ing to pay for a click. In our opin­ion, CPCs are a bet­ter mea­sure of the mar­ket because they reflect what peo­ple are ACTUALLY pay­ing Face­book. Sug­gested CPCs are at best an infer­en­tial measure.

Sec­ond, we com­pared data from Q4 2010 to Q1 2011 while TBGs num­bers trend from March 19th to April 11th, a period of about three weeks so it is a much shorter time frame in which to draw conclusions.

Third, the method­ol­ogy of the analy­ses has not been men­tioned. I would like to point out that for Effi­cient Frontier’s analy­sis, we exam­ined a broad range of clients from a vari­ety of sec­tors includ­ing retail, finan­cial ser­vices, travel, and auto­mo­tive.  Our index includes more than 10 enter­prise class cus­tomers that have been Effi­cient Fron­tier clients on Face­book for more than 2 con­sec­u­tive quar­ters. While we feel that our dataset is fairly robust and direc­tion­ally cor­rect, we are aware that there would be some vari­ance in the num­bers. This is typ­i­cal of an early stage, rapidly grow­ing chan­nel where the mar­ket­place is very dynamic. How­ever, a +/-40% dis­crep­ancy is star­tling. Which brings us to the crux of the debate. Which num­ber reports the cor­rect trend?

Math­e­mat­i­cally CPCs, click vol­ume and Facebook’s rev­enue from Mar­ket­place ads have a very sim­ple rela­tion­ship. Facebook’s mar­ket­place rev­enue is the prod­uct of the CPCs that Face­book charges adver­tis­ers and the clicks that adver­tis­ers get .

If CPCs are truly down 30% the per­cent­age of paid clicks in a quar­ter would have to increase by 30% for Fsacebook’s rev­enue to remain flat. Assum­ing that audi­ence reach is rel­a­tively sta­ble, the ad serv­ing pop­u­la­tion would have to almost triple in a year in the U.S. for Face­book to record the same rev­enue as last year. Anec­do­tal evi­dence shows that such an increase in inven­tory is unlikely. As one com­men­tor on All­Face­book points out, the roll out of the new pro­file in Jan­u­ary would have actu­ally decreased inven­tory as the “More Ads” link was removed.

This is con­trary to all the trends and reports that pre­dict that Face­book will dou­ble rev­enues in 2011. A look at the early days of search sug­gests that there will be some of both CPC increases and paid click increases going on. On the demand side, adver­tis­ers are rapidly lever­ag­ing the plat­form and are will­ing to pay more for their ads while on the sup­ply side, Face­book is grow­ing its audi­ence reach so that there is an ever greater num­ber of ads that Face­book can serve its ads. How­ever, to record a dou­bling of rev­enues we believe that the bulk of the growth will come from the increase in CPCs at this stage.

In sum­mary, there are 3 rea­sons why we think our num­bers are direc­tion­ally cor­rect. First and fore­most, we report on the actual CPC met­ric and not an infer­en­tial, indi­rect num­ber that does not nec­es­sar­ily reflect the mar­ket. Sec­ond, we have a robust method­ol­ogy that looks at longer term data. And finally, our num­bers are con­sis­tent with the anec­do­tal evi­dence and reports about Facebook’s growth in 2011.

I hope this helps clear up the discussion.

Dr. Sid­dharth Shah
Sr. Direc­tor, Busi­ness Analytics